What is tail risk?

Tail risk is the risk of a rare and extreme event, one that sits in the far ends, or tails, of the range of possible outcomes. These events happen seldom, but when they do they can cause losses much larger than anything seen in normal times. It matters because averages and typical-case planning do not show them.

How it works

If outcomes followed a bell curve, very large moves would be almost impossible. In many real systems, such as financial markets, supply chains and power grids, extreme events happen more often than a bell curve suggests. This is called a fat tail. Shocks also tend to be linked: when one thing fails, others fail at the same time, which makes the extreme outcome worse.

The practical trade-off is cost against protection. Insurance, reserves and diversified suppliers cost money every period and may seem wasted in the many years when nothing happens. But they are what keeps the system alive in the one year when the extreme event arrives, so the judgement is about survival, not average profit.

Example

Say a business earns a profit of 100,000 in a normal year. Once in 50 years, a disruption causes a loss of 3,000,000. Over 50 years, normal profit adds up to about 4,900,000 from the 49 good years, and the single loss wipes out most of it, leaving 1,900,000. If the company had only 500,000 in reserves when it struck, it would fail before ever seeing the average. This is an illustration with round numbers.

How to model it in Machinations

Create a profit Source with a small random variation per step, and a Gate that with a small Lever-set probability sends a large amount to a Drain on the cash Pool. Add an End condition for when the Pool goes below zero. Play many times at once, for example thousands, and count how often the Pool ends up empty. Increase reserves via a Lever and see how the failure count drops.